INDEX

Explanations

words related to critical evaluation or assessment

oai_token-act-pair · gpt-3.5-turbo

terms associated with quality and distress signals

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

otomy

-0.57

 Dragonbound

-0.57

ships

-0.57

eous

-0.56

itarian

-0.53

ayson

-0.53

iller

-0.53

asar

-0.52

arus

-0.52

ship

-0.52

POSITIVE LOGITS

ãĥŁ

0.70

med

0.68

rings

0.67

%%%%

0.66

TER

0.65

spect

0.64

âĸ¬

0.64

Ĳ

0.64

mm

0.64

hed

0.63

Activations Density 0.221%

words related to critical evaluation or assessment

terms associated with quality and distress signals

No Comments

No Known Activations